ChatterBank0 min ago
Yougov Now Says Hung Parliament
Blimey
Answers
Best Answer
No best answer has yet been selected by scooping. Once a best answer has been selected, it will be shown here.
For more on marking an answer as the "Best Answer", please visit our FAQ.
Gawd help us, I can believe mikey is still banging on about how Hilary really won the US election. Last time I checked me old china she was not the president. Anyway TBH I am worried about the Tories campaign they do seem to be shooting wide of a lot of open goals at the moment so there may be something in the Poll picture.
But what do you mean by wrong?
If I tell you that, say, the Tories will win between 31% and 45% of the vote in a seat, and Labour between 39% and 52% of the vote in that same seat (both results with 95% confidence), then I expect that the seat will go to Labour -- but, also, it could go to the Tories within the errors I quoted. So, if it does end up going to the Tories (by, say, 44 to 41), is the prediction wrong?
If I tell you that, say, the Tories will win between 31% and 45% of the vote in a seat, and Labour between 39% and 52% of the vote in that same seat (both results with 95% confidence), then I expect that the seat will go to Labour -- but, also, it could go to the Tories within the errors I quoted. So, if it does end up going to the Tories (by, say, 44 to 41), is the prediction wrong?
What is the problem with a Hung Parliament?
An argument could be made that it would be the best of both worlds.
Both main parties putting forward their best candidates for the posts.
At the least it should ensure Diane Abbott is given a job like
'Overseer of maypole dancing rhythms' that will keep her out of trouble.
An argument could be made that it would be the best of both worlds.
Both main parties putting forward their best candidates for the posts.
At the least it should ensure Diane Abbott is given a job like
'Overseer of maypole dancing rhythms' that will keep her out of trouble.
In that case aren't you essentially requiring what ... well, nearly every scientist would regard as impossible standards? It stands to reason that you can't say, to the third decimal point or whatever, what vote share each party will get. All you can do is estimate, to within certain confidence intervals, and give an indication of which result appears to be more likely.
If you are requiring pinpoint accuracy in your definition of "accurate" -- as in, say, 45.38% is "wrong" if the actual result is 45.4% -- then no wonder you're dismissing these results. I'm not sure I'm aware of anything experimental or statistical that's ever been done that would meet your required standards.
As long as you are consistent -- and that's what I've seen so far -- then I guess that's fine to have such standards. But at the same time, it's not unreasonable to have less exacting, more statistically-minded standards for what counts as accurate. Short of holding a general election, you can't measure any more accurately than this sort of technique. And the thing is, it does work. Central values reported by polls often end up missing, but that's to be expected. What matters is the error either side -- and it's relatively rare for the result to fall outside the expected confidence interval.
If you are requiring pinpoint accuracy in your definition of "accurate" -- as in, say, 45.38% is "wrong" if the actual result is 45.4% -- then no wonder you're dismissing these results. I'm not sure I'm aware of anything experimental or statistical that's ever been done that would meet your required standards.
As long as you are consistent -- and that's what I've seen so far -- then I guess that's fine to have such standards. But at the same time, it's not unreasonable to have less exacting, more statistically-minded standards for what counts as accurate. Short of holding a general election, you can't measure any more accurately than this sort of technique. And the thing is, it does work. Central values reported by polls often end up missing, but that's to be expected. What matters is the error either side -- and it's relatively rare for the result to fall outside the expected confidence interval.
Jim, Science isn’t in question. I accept that science gets it wrong at times. I know you have a fascination for figures (of the numerical kind), but I don’t, and in the run up to a General Election I think it’s utterly wrong to publish what could be a grossly inaccurate prediction that may well sway sections of the electorate one way or the other. You said yourself there hasn't been a poll since April last year that had Labour ahead (and only three total since the 2015 election), and yet here we have a headline predicting that there will be a hung parliament. Sorry to disappoint you but I’m not likely to change my mind. Can we please put this to bed now?
"Can we please put this to bed now?"
I'd rather not. Discussions aren't always about changing someone's mind. They can be interesting in themselves; or perhaps someone reading this might find it interesting even if you, the participant, do not.
I explained already why the hung parliament could be correct even with Labour behind, as long as they are close enough. If I'm trying to persuade you of anything else it's that this poll isn't as wrong as you are treating it -- but that doesn't make the central "hung parliament" figures right either, necessarily. I'm trying to make the point about the bands of predictions, that apply to each seat and so to the prediction as a whole. That adds uncertainty without making the whole exercise pointless, because you can understand and quantify that uncertainty.
"I think it’s utterly wrong to publish what could be a grossly inaccurate prediction that may well sway sections of the electorate one way or the other."
Well, possibly. I think the problem is more in how they are reported, rather than whether they are reported at all. The subtlety you are bored of is vital but so often ignored -- which cuts both ways. People think polls are wrong because all they hear are the headline figures that are usually wrong (because, well, obviously); other people get too excited about one particular poll because they fail to appreciate that it could be an outlier.
If all that were publicised more widely then there'd be no problem. It isn't, so there is a little.
I had tried to wrap this up earlier actually -- maybe it was a timid effort as I'm too interested in polling and electoral analysis these days -- but you can perhaps draw some comfort from this poll. Even if it ends up being wrong, it could partly be wrong because enough wavering Tory voters took it seriously and went out to vote accordingly.
Next week will tell us how accurate (or not) this particular poll was, but as long as it's put in proper context there's no harm in publishing it.
I'd rather not. Discussions aren't always about changing someone's mind. They can be interesting in themselves; or perhaps someone reading this might find it interesting even if you, the participant, do not.
I explained already why the hung parliament could be correct even with Labour behind, as long as they are close enough. If I'm trying to persuade you of anything else it's that this poll isn't as wrong as you are treating it -- but that doesn't make the central "hung parliament" figures right either, necessarily. I'm trying to make the point about the bands of predictions, that apply to each seat and so to the prediction as a whole. That adds uncertainty without making the whole exercise pointless, because you can understand and quantify that uncertainty.
"I think it’s utterly wrong to publish what could be a grossly inaccurate prediction that may well sway sections of the electorate one way or the other."
Well, possibly. I think the problem is more in how they are reported, rather than whether they are reported at all. The subtlety you are bored of is vital but so often ignored -- which cuts both ways. People think polls are wrong because all they hear are the headline figures that are usually wrong (because, well, obviously); other people get too excited about one particular poll because they fail to appreciate that it could be an outlier.
If all that were publicised more widely then there'd be no problem. It isn't, so there is a little.
I had tried to wrap this up earlier actually -- maybe it was a timid effort as I'm too interested in polling and electoral analysis these days -- but you can perhaps draw some comfort from this poll. Even if it ends up being wrong, it could partly be wrong because enough wavering Tory voters took it seriously and went out to vote accordingly.
Next week will tell us how accurate (or not) this particular poll was, but as long as it's put in proper context there's no harm in publishing it.
Jim, love, I'm not joining in 'cos my eyes glazed reading through this post - no way am I qualified to spout - but can you realise, please, that most of us don't think solely scientifically? I'm sure stats. influence you- but not people like me for instance (OK I'M weird - but there are a lot like me).
"What is the problem with a Hung Parliament?"
Good grief, Eddie, if you need to ask that just take a look at nations that have voting systems that result in "hung" parliaments of various descriptions. In particular look at Italy where the average lifespan of governments since Mussolini is a little over 12 months. Nothing gets done; no critical decisions are taken; everybody gets what nobody wants. In fact, everybody gets next to nothing from government because politicians are too busy jockeying for position and preparing for the next election. Better still, have a look at government in the UK from 2010 to 2015.
A hung Parliament is the last thing we should wish for (even behind a Labour government, and that's saying something coming from me).
Good grief, Eddie, if you need to ask that just take a look at nations that have voting systems that result in "hung" parliaments of various descriptions. In particular look at Italy where the average lifespan of governments since Mussolini is a little over 12 months. Nothing gets done; no critical decisions are taken; everybody gets what nobody wants. In fact, everybody gets next to nothing from government because politicians are too busy jockeying for position and preparing for the next election. Better still, have a look at government in the UK from 2010 to 2015.
A hung Parliament is the last thing we should wish for (even behind a Labour government, and that's saying something coming from me).
Latest Posts
Quizzes & Puzzles1 min ago
Crosswords2 mins ago
Quizzes & Puzzles3 mins ago
Quizzes & Puzzles11 mins ago
Will You Be Shopping At Boots This...
News15 mins ago
Related Questions
Sorry, we can't find any related questions. Try using the search bar at the top of the page to search for some keywords, or choose a topic and submit your own question.